DIY Tesla Drive Unit Repair

[David Harvey]

In the meantime, I'm heading to Europe Monday to set up EU distribution. My wife thinks it is a vacation. Seriously though, I have some family that will get a supply which will significantly help shipping and taxes inside the EU. In the short term, message me here or on FB to discuss. Eventually, he should have something more formal set up.

I don't suppose you are coming to the UK by any chance?

 

[vanR]

Datapoint update: The Ceimin triple lip Teflon seal still seems to properly seal in my LDU after 10563mi = 17000km. See also:

 

[Hank42]

Nice! I'd still perform the coolant delete mod.. "just in case"

 

[howardc64]

Datapoint update: The Ceimin triple lip Teflon seal still seems to properly seal in my LDU after 10563mi = 17000km. See also:

Impressive!

I would say have a lot to do with condition of the shaft. @vanR's rotor shaft was nearly pristine from an original LDU motor (not reman) Had a single lip that leaked (seems like all Rav4 EV's LDU had single lip installed. Rav4EV probably sourced LDU when Tesla already made the switch to single lip) and caught early.

Just checked @mr_hyde 2014 original LDU last week at 80k+ miles. Bone dry (wiped clean the grease from last 2 checks) as well.

It does suggest new LDU shaft + triple lip PTFE will last (And all were installed with ample amount of grease. some spun off onto the speed sensor visible on the first pull for inspection). Unfortunately for Model S, most original LDU were removed due to clunk+milling noise problem and single lipped reman got installed. Rav4EV seems to all get single lip on new LDUs.

 

[Concreteseller]

hello all, I have a 12-14 MS that had the noisy ldu replaced under warranty about 2 years and 40k miles ago and now apparently has the seal problem. Not enough posts to reach out to howards or anyone for that matter but i'd like to diy and have some thoughts, nothing novel, but would like to pick a brain or 2. quote from tesla was 7200 which sounds about right.

 

[howardc64]

hello all, I have a 12-14 MS that had the noisy ldu replaced under warranty about 2 years and 40k miles ago and now apparently has the seal problem. Not enough posts to reach out to howards or anyone for that matter but i'd like to diy and have some thoughts, nothing novel, but would like to pick a brain or 2. quote from tesla was 7200 which sounds about right.

@vanR maintained website

TeslaLDU

Table of Contents

sites.google.com

 

[ajbessinger]

Impressive!

I would say have a lot to do with condition of the shaft. @vanR's rotor shaft was nearly pristine from an original LDU motor (not reman) Had a single lip that leaked (seems like all Rav4 EV's LDU had single lip installed. Rav4EV probably sourced LDU when Tesla already made the switch to single lip) and caught early.

Just checked @mr_hyde 2014 original LDU last week at 80k+ miles. Bone dry (wiped clean the grease from last 2 checks) as well.

It does suggest new LDU shaft + triple lip PTFE will last (And all were installed with ample amount of grease. some spun off onto the speed sensor visible on the first pull for inspection). Unfortunately for Model S, most original LDU were removed due to clunk+milling noise problem and single lipped reman got installed. Rav4EV seems to all get single lip on new LDUs.

Pretty much all RAV4s originally came with triple lip seals. Much like the Model S, a very large portion of RAV4 drive units were replaced early on for noise issues and rotor spline wear, and any replaced after ~late-2015 have the single lip seals (just like Model S/X).

In fact, I believe the RAV4 EV may have actually hit dealer lots before the first Model S deliveries started. Even current RAV4 LDUs actually carry a lot of the hallmark features of the very early style LDU, including the old style supply side bus bar setup with the tiny screws attaching to the voltage sense board, which is attached to one of the phases with standoffs. Earlier builds even had the extra "air bleed" tube that runs from the coolant outlet to the top of the gearbox that is only present on the very earliest of LDUs.

I have actually have an original, very early RAV4 LDU in my shop right now, which was built in Oct of 2012 (seal is marked 09/25/12), and has a serial number of only 2292 (RAV4 and Mercedes LDUs are serialized in parallel with Model S/X LDUs). I'll attach some pics of it to show everything that I'm talking about.

 

[howardc64]

Pretty much all RAV4s originally came with triple lip seals. Much like the Model S, a very large portion of RAV4 drive units were replaced early on for noise issues and rotor spline wear, and any replaced after ~late-2015 have the single lip seals (just like Model S/X).

Good to know, are these the stats you see?

• < 2015 (LDU date) triple lip?
• >= ~late 2015 onwards LDU (new or reman) are single lip

 

[ajbessinger]

Good to know, are these the stats you see?

• < 2015 (LDU date) triple lip?
• >= ~late 2015 onwards LDU (new or reman) are single lip

More or less, I've never recorded exact build dates, but based on what I've seen, the cutoff for the end of use of triple lip seals seems to be near the end of 2015 or maybe very early 2016. So there's sort of a "goldilocks" period starting from late 2014 where the motors got the updated rotor, ceramic bearings, etc., in addition to retaining the old triple lip seal.
Of course, when it comes to reman motors, the serial number on the sticker is updated (LDUs are now well over 1M built/reman'd), while the original build date always stays the same. For example, I've seen lots of R1 and R2 LDUs with an O.G. build date of 2013, but serial numbers well into the high hundreds of thousands (and one or two that were 1M+).

 

[howardc64]

More or less, I've never recorded exact build dates, but based on what I've seen, the cutoff for the end of use of triple lip seals seems to be near the end of 2015 or maybe very early 2016. So there's sort of a "goldilocks" period starting from late 2014 where the motors got the updated rotor, ceramic bearings, etc., in addition to retaining the old triple lip seal.
Of course, when it comes to reman motors, the serial number on the sticker is updated (LDUs are now well over 1M built/reman'd), while the original build date always stays the same. For example, I've seen lots of R1 and R2 LDUs with an O.G. build date of 2013, but serial numbers well into the high hundreds of thousands (and one or two that were 1M+).

haha yeah, a few goldilocks LDUs exist (@mr_hyde has one, 2014 original 80k+ miles, bone dry on recent 3nd inspection). Some owner even swear LDUs are bullet proof after 10 years 150k miles. Probably because they got the goldilocks config. But seems quite rare due to SCs aggressively swapping out LDUs with remans all the way through 2017+. ~2018 no one at SCs could heard any more LDU noises haha. My impression of this sudden behavior shift = HQ said "stop".

 

[meklaren]

Dear Forum Members,

First of all, I would like to thank the forum and the YouTube videos (especially DIYEVGUY, DIY EV Guy and https://sites.google.com/view/teslaldu/ and other websites, detailed comments) for the wealth of information. I own a beloved 2014 Model S P85.

My long story:
It started when I was heading to Obertauern in Austria, transitioning from 16°C to -4°C in the mountains (yes, I was going skiing). I stopped by the roadside, and my car displayed a message: "vehicle may not start" along with error codes BMS_w035 and BMS_f035. I was terrified I’d be stranded in the cold, but I managed to reach the nearest Supercharger station. Since my car is out of warranty, I sought help elsewhere.
A Facebook group suggested contacting a Tesla employee (whose name I won’t mention) who promised to check the car within two hours, though a third party only looked at it remotely two days later. Meanwhile, I awaited guidance on what to do with my poor car. They remotely diagnosed motor water ingress. The employee repeatedly advised against disassembling it myself, citing difficulty despite having several mechanical engineer friends. I recognized this as typical Tesla employee marketing. I guess, being an insider, he had lots of internal information. TS quoted me an estimated repair cost of 7,000 EUR (almost the same in USD), likely for a motor replacement. Afterward, no errors reappeared, and I spent 1 day skiing and after I could drive home, covering 500 kilometers.
At home I asked third-party companies, quoted at least ~1,300 EUR, potentially up to 3,500 EUR, depending on the extent of the issue, offering to replace seals, bearings, and other parts. However, seal failure could recur later. I found other companies willing to refurbish the motor for 3,000 EUR, likely offering quality service.
We decided to attempt the repair ourselves since we had the capability. Fortunately, my family is supportive; my cousin owns an auto service shop, and my brother is licensed to work on electric cars. Together, we followed the service manual and utilized numerous YouTube videos and forum tips. We chose to eliminate the rotor cooling system, trusting it would work well under normal driving conditions, especially since Tesla sells motors with coolant delete in the -U version. My motor was a 1025598-00-P version, replaced in 2017 and having run 130.000 km (80.778 miles) The main battery was replaced in a Barcelona service center in 2019. We also used 3D printer for prototype measures.
Upon disassembly, we realized the error codes might not have been caused by water ingress. Just little rust showed. The stator, rotor, and bearings didn’t show signs of very rusty water damage. Only the inverter side cover had some fluid, but we couldn’t trace its origin, likely residual from a previous refurbishment. Disassembling the motor was preventive and enriching from a technical standpoint. The outer manifold bearing was little bit in bad condition. Removing the rotor was difficult (unlike some easy YouTube videos), requiring us to machine an extra shaft collar.
We decided to invest in tools rather than paying to others (if you have courage and time, recommended this way). I can confirm that working on a Tesla is indeed easier than on ICE car. Despite initial challenges, we overcame them, ensuring optimal performance with proper parts and research. Although I lack programming skills, I would appreciate access to the Developer menu to monitor temperatures. In person, Tesla service was unhelpful, withholding information about the coolant delete.

I am pleased to share the manifold we designed and manufactured. I have been using the prototype for a while, covering 1000 km without any significant issues (outside temperature 20-30 degrees= 68-86 Fahrenheit). We are offering this manifold for sale. If anyone is interested, please contact me at [email protected]. The cost is 450 USD / 450 EUR + shipping. We can issue invoices.

An installation video is coming soon.Meanwhile, I’ve attached some pictures. I hope my story helps others maintain their Teslas for a long time.

Purchased/needed parts:
- Axle seal
- 1065932-00-A
- Mannol G11 5l concentrate 50:50
- Bearing SKF 6007-2RZTN9/HC5C3WT
- Dexron VI oil (Motul ATF 6) 2 liters
- Loctite 243 thread locker
- Loctite 648 bearing retaining
- Würth brake cleaner
- CX80 silicone sealant
- Car cover
- Brake fluid DOT4 - Textar
- Contact spray
- Custom manifold manufacturing ALU 5083.
-many tools (like torque wrench, bearing puller, machine press etc)

Additionally, we greased the axle shafts, checked the brake fluid, inspected the brakes, and lubricated the suspension leveler joints. Upon reassembly, the car initially wouldn’t start without any error message. The solution was to disconnect the 12V negative terminal and fireloop again, wait, and reconnect it. Bleeding the system was easier following a previous comment by connecting the charger in and out.
When attempting to charge at the SC, a BMS_w035 error code appeared once, but nothing since. Tesla service described it as a generic code, possibly caused by various issues like a broken cable. After a system check, everything was fine. Online sources suggest it could be a fuse problem, battery heater issue, or worst-case scenario, the battery itself. Interestingly, the main battery showed 30 more kms range after 5 weeks of inactivity.

I am unsure about the specific issue related to the f_035 error code. The entire repair and design process took 31 hours, being our first major repair, but the complete fix, including parts waiting, took 5 weeks.

I’m delighted the car is running again, and I look forward to many more kilometers despite minor issues.
So again, if anyone wants to order a manifold, please email [email protected]. The manifold price is 450 USD / 450 EUR plus shipping.
All the best

I can attach only 10 pictures so I will upload more in the next post.

As I wrote before in that post, finally we made that short video.

If anyone interested, here is the youtube link:

Tesla LDU coolant delete



(I hope moderators will approve my posts soon.)

 

[brainhouston]

nice, appreciate the quick video, good editing skills! (so tired of everyone making useless long videos for no reason lol)

seems that even with rotor cooling, surface gets that hotter in the middle signature?..

 

[ajbessinger]

haha yeah, a few goldilocks LDUs exist (@mr_hyde has one, 2014 original 80k+ miles, bone dry on recent 3nd inspection). Some owner even swear LDUs are bullet proof after 10 years 150k miles. Probably because they got the goldilocks config. But seems quite rare due to SCs aggressively swapping out LDUs with remans all the way through 2017+. ~2018 no one at SCs could heard any more LDU noises haha. My impression of this sudden behavior shift = HQ said "stop".

I just finished up work on a "goldilocks" config, probably one of the highest mileage original, untouched LDUs I've personally seen. Sept 2015 build S85 with ~233k mi on the original LDU (and of course original seal, which was starting to leak). Amazingly, it had pretty much zero bearing noise.
It of course got rotor coolant delete, and also needed a new inverter cover and cables, as they were severely corroded and allowing moisture intrusion into the inverter cavity. The owner got VERY lucky that his inverter didn't get destroyed by the water intrusion.